基于准时化生产要求下改进型粒子群算法研究
|
摘要
针对新形势下制造型企业对准时化生产的要求,运用人工智能算法可以实现生产过程中对多工位的准时化配送,如粒子群算法,但是单一的粒子群算法在应用的时候容易出现"早熟"现象,且收敛精度差。为此设计了一种改进型粒子群算法,引入遗传算法的编码、交叉变异策略和模拟退火算法,三者相融合,增强算法收敛速度和精度的同时又抑制算法"早熟"从而得到全局最优。同时企业可根据实际生产需求,通过改变适应函数的惯性权重来安排生产。最后通过实例验证该算法的有效性。
In response to the requirements of JIT(Just in time) from manufacturing enterprises in the new situation, using artificial intelligence algorithm, such as particle swarm optimization, can realize the punctual delivery of multiple stations in the production process. However the single particle swarm optimization algorithm is prone to "premature" phenomenon and has poor convergence accuracy. So an improved PSO algorithm is designed for this purpose. The genetic algorithm coding, crossover mutation strategy and simulated annealing algorithm are introduced and merged. This can enhance the convergence speed and accuracy of the algorithm while suppressing the algorithm "premature" to get the global optimum. At the same time, according to the actual production needs, the inertia weight of the adaptive function can be changed to schedule production. Finally, an experimental case is given to verify the effectiveness of the proposed algorithm.
In response to the requirements of JIT(Just in time) from manufacturing enterprises in the new situation, using artificial intelligence algorithm, such as particle swarm optimization, can realize the punctual delivery of multiple stations in the production process. However the single particle swarm optimization algorithm is prone to "premature" phenomenon and has poor convergence accuracy. So an improved PSO algorithm is designed for this purpose. The genetic algorithm coding, crossover mutation strategy and simulated annealing algorithm are introduced and merged. This can enhance the convergence speed and accuracy of the algorithm while suppressing the algorithm "premature" to get the global optimum. At the same time, according to the actual production needs, the inertia weight of the adaptive function can be changed to schedule production. Finally, an experimental case is given to verify the effectiveness of the proposed algorithm.
引文
[1] 曹健,张申生,牟玉洁,等.面向订单生产方式的产品与过程集成配置技术[J].中国机械工程,2004,15(19):1710-1712.
[2] 蒋丽,丁斌,减晓宁.以工位为中心的生产物流配送优化[J].计算机集成制造系统,2009,15(11):2154-2159.
[3] 鲁建厦,朱恺,董巧英.基于混合混沌粒子群算法的装配线平衡问题研究[J].浙江工业大学学报,2017,45(1):114-118.
[4] GOLZ J,GUJJULA R,GüNTHER H O,et al.Part feeding at high-variant mixed-model assembly lines[J].Flexible services and manufacturing journal,2012,24(2):119-141.
[5] 党立伟,孙小明.基于配送BOM的装配线物料循环准时化配送研究[J].机械制造,2012,50(7):89-91.
[6] 沈维蕾,马鹏辉.基于MES的混流装配线物料配送方法研究[J].合肥工业大学学报(自然科学版),2013(10):1166-1169.
[7] 鲁建厦,朱恺,董巧英.基于混合混沌粒子群算法的装配线平衡问题研究[J].浙江工业大学学报,2017,45(1):114-118.
[8] GLURE D.Optimization of automated material handling systems regarding robustness and delivery time[J].IEEE international symposium on semiconductor manufacturing conference and proceedings,2003(1):96-100.
[9] 陈志杨,刘妍.改进粒子群搜索的二维皮革排样优化算法[J].浙江工业大学学报.2015,43(5):492-496.
[10] 敖永才,师奕兵,张伟,等.自适应惯性权重的改进粒子群算法[J].电子科技大学学报,2014,43(6):874-880.
[11] 张则强,余庆良,胡俊逸,等.随机混流装配线平衡问题的一种混合粒子群算法[J].机械设计与研究,2013,29(2):60-63.
[2] 蒋丽,丁斌,减晓宁.以工位为中心的生产物流配送优化[J].计算机集成制造系统,2009,15(11):2154-2159.
[3] 鲁建厦,朱恺,董巧英.基于混合混沌粒子群算法的装配线平衡问题研究[J].浙江工业大学学报,2017,45(1):114-118.
[4] GOLZ J,GUJJULA R,GüNTHER H O,et al.Part feeding at high-variant mixed-model assembly lines[J].Flexible services and manufacturing journal,2012,24(2):119-141.
[5] 党立伟,孙小明.基于配送BOM的装配线物料循环准时化配送研究[J].机械制造,2012,50(7):89-91.
[6] 沈维蕾,马鹏辉.基于MES的混流装配线物料配送方法研究[J].合肥工业大学学报(自然科学版),2013(10):1166-1169.
[7] 鲁建厦,朱恺,董巧英.基于混合混沌粒子群算法的装配线平衡问题研究[J].浙江工业大学学报,2017,45(1):114-118.
[8] GLURE D.Optimization of automated material handling systems regarding robustness and delivery time[J].IEEE international symposium on semiconductor manufacturing conference and proceedings,2003(1):96-100.
[9] 陈志杨,刘妍.改进粒子群搜索的二维皮革排样优化算法[J].浙江工业大学学报.2015,43(5):492-496.
[10] 敖永才,师奕兵,张伟,等.自适应惯性权重的改进粒子群算法[J].电子科技大学学报,2014,43(6):874-880.
[11] 张则强,余庆良,胡俊逸,等.随机混流装配线平衡问题的一种混合粒子群算法[J].机械设计与研究,2013,29(2):60-63.